- Predicting 30-day mortality in severely injured elderly patients with trauma in Korea using machine learning algorithms: a retrospective study
-
Jonghee Han, Su Young Yoon, Junepill Seok, Jin Young Lee, Jin Suk Lee, Jin Bong Ye, Younghoon Sul, Se Heon Kim, Hong Rye Kim
-
J Trauma Inj. 2024;37(3):201-208. Published online August 8, 2024
-
DOI: https://doi.org/10.20408/jti.2024.0024
-
-
Abstract
PDF
- Purpose
The number of elderly patients with trauma is increasing; therefore, precise models are necessary to estimate the mortality risk of elderly patients with trauma for informed clinical decision-making. This study aimed to develop machine learning based predictive models that predict 30-day mortality in severely injured elderly patients with trauma and to compare the predictive performance of various machine learning models. Methods: This study targeted patients aged ≥65 years with an Injury Severity Score of ≥15 who visited the regional trauma center at Chungbuk National University Hospital between 2016 and 2022. Four machine learning models—logistic regression, decision tree, random forest, and eXtreme Gradient Boosting (XGBoost)—were developed to predict 30-day mortality. The models’ performance was compared using metrics such as area under the receiver operating characteristic curve (AUC), accuracy, precision, recall, specificity, F1 score, as well as Shapley Additive Explanations (SHAP) values and learning curves. Results: The performance evaluation of the machine learning models for predicting mortality in severely injured elderly patients with trauma showed AUC values for logistic regression, decision tree, random forest, and XGBoost of 0.938, 0.863, 0.919, and 0.934, respectively. Among the four models, XGBoost demonstrated superior accuracy, precision, recall, specificity, and F1 score of 0.91, 0.72, 0.86, 0.92, and 0.78, respectively. Analysis of important features of XGBoost using SHAP revealed associations such as a high Glasgow Coma Scale negatively impacting mortality probability, while higher counts of transfused red blood cells were positively correlated with mortality probability. The learning curves indicated increased generalization and robustness as training examples increased. Conclusions: We showed that machine learning models, especially XGBoost, can be used to predict 30-day mortality in severely injured elderly patients with trauma. Prognostic tools utilizing these models are helpful for physicians to evaluate the risk of mortality in elderly patients with severe trauma.
-
Summary
- Validation of chest trauma scoring systems in polytrauma: a retrospective study with 1,038 patients in Korea
-
Hongrye Kim, Mou Seop Lee, Su Young Yoon, Jonghee Han, Jin Young Lee, Junepill Seok
-
J Trauma Inj. 2024;37(2):114-123. Published online May 9, 2024
-
DOI: https://doi.org/10.20408/jti.2023.0087
-
-
2,076
View
-
90
Download
-
1
Citations
-
Abstract
PDFSupplementary Material
- Purpose
Appropriate scoring systems can help classify and treat polytrauma patients. This study aimed to validate chest trauma scoring systems in polytrauma patients.
Methods Data from 1,038 polytrauma patients were analyzed. The primary outcomes were one or more complications: pneumonia, chest complications requiring surgery, and mortality. The Thoracic Trauma Severity Score (TTSS), Chest Trauma Score, Rib Fracture Score, and RibScore were compared using receiver operating characteristic (ROC) analysis in patients with or without head trauma.
Results In total, 1,038 patients were divided into two groups: those with complications (822 patients, 79.2%) and those with no complications (216 patients, 20.8%). Sex and body mass index did not significantly differ between the groups. However, age was higher in the complications group (64.1±17.5 years vs. 54.9±17.6 years, P<0.001). The proportion of head trauma patients was higher (58.3% vs. 24.6%, P<0.001) and the Glasgow Coma Scale score was worse (median [interquartile range], 12 [6.5–15] vs. 15 [14–15]; P<0.001) in the complications group. The number of rib fractures, the degree of rib fracture displacement, and the severity of pulmonary contusions were also higher in the complications group. In the area under the ROC curve analysis, the TTSS showed the highest predictive value for the entire group (0.731), head trauma group (0.715), and no head trauma group (0.730), while RibScore had the poorest performance (0.643, 0.622, and 0.622, respectively)
Conclusions Early injury severity detection and grading are crucial for patients with blunt chest trauma. The chest trauma scoring systems introduced to date, including the TTSS, are not acceptable for clinical use, especially in polytrauma patients with traumatic brain injury. Therefore, further revisions and analyses of chest trauma scoring systems are recommended.
-
Summary
-
Citations
Citations to this article as recorded by
- Clinical prediction models for the management of blunt chest trauma in the emergency department: a systematic review
Ceri Battle, Elaine Cole, Kym Carter, Edward Baker BMC Emergency Medicine.2024;[Epub] CrossRef
- Clinical characteristics and mortality risk factors among trauma patients by age groups at a single center in Korea over 7 years: a retrospective study
-
Jonghee Han, Su Young Yoon, Junepill Seok, Jin Young Lee, Jin Suk Lee, Jin Bong Ye, Younghoon Sul, Seheon Kim, Hong Rye Kim
-
J Trauma Inj. 2023;36(4):329-336. Published online November 7, 2023
-
DOI: https://doi.org/10.20408/jti.2023.0035
-
-
Abstract
PDF
- Purpose
In this study, we aimed to compare the characteristics of patients with trauma by age group in a single center in Korea to identify the clinical characteristics and analyze the risk factors affecting mortality.
Methods Patients aged ≥18 years who visited the Chungbuk National University Hospital Regional Trauma Center between January 2016 and December 2022 were included. The accident mechanism, severity of the injury, and outcomes were compared by classifying the patients into group A (18–64 years), group B (65–79 years), and group C (≥80 years). In addition, logistic regression analysis was performed to identify factors affecting death.
Results The most common injury mechanism was traffic accidents in group A (40.9%) and slipping in group B (37.0%) and group C (56.2%). Although group A had the highest intensive care unit admission rate (38.0%), group C had the highest mortality rate (9.5%). In the regression analysis, 3 to 8 points on the Glasgow Coma Scale had the highest odds ratio for mortality, and red blood cell transfusion within 24 hours, intensive care unit admission, age, and Injury Severity Score were the predictors of death.
Conclusions For patients with trauma, the mechanism, injured body region, and severity of injury differed among the age groups. The high mortality rate of elderly patients suggests the need for different treatment approaches for trauma patients according to age. Identifying factors affecting clinical patterns and mortality according to age groups can help improve the prognosis of trauma patients in the future.
-
Summary
- Surgical management of supratentorial and infratentorial epidural hematoma in Korea: three case reports
-
Su Young Yoon, Junepill Seok, Yook Kim, Jin Suk Lee, Jin Young Lee, Mou Seop Lee, Hong Rye Kim
-
J Trauma Inj. 2023;36(4):399-403. Published online December 26, 2023
-
DOI: https://doi.org/10.20408/jti.2023.0073
-
-
Abstract
PDF
- Supratentorial and infratentorial epidural hematoma (SIEDH) is a rare but life-threatening complication following traumatic brain injury. However, the literature on SIEDH is sparse, consisting only of a few small series. Prompt diagnosis and the application of appropriate surgical techniques are crucial for the rapid and safe management of SIEDH. Herein, we present three cases of SIEDH treated at our institution, employing a range of surgical approaches.
-
Summary
- Chronic Traumatic Glass Foreign Body Removal from the Lung through a Direct Parenchymal Incision
-
Su Young Yoon, Si Wook Kim, Jin Suk Lee, Jin Young Lee, Jin Bong Ye, Se Heon Kim, Young Hoon Sul
-
J Trauma Inj. 2019;32(4):248-251. Published online December 30, 2019
-
DOI: https://doi.org/10.20408/jti.2019.031
-
-
4,581
View
-
63
Download
-
1
Citations
-
Abstract
PDF
Traumatic intrapulmonary glass foreign bodies that are missed on an initial examination can migrate and lead to severe complications. Here, we present a rare case of a traumatic intrapulmonary glass foreign body surgically removed by a direct pulmonary incision, which preserved the pulmonary parenchyma and avoided severe complications caused by migration.
-
Summary
-
Citations
Citations to this article as recorded by
- Thoracoscopic retrieval of an intrapulmonary sewing needle: A case report
Houssem Messaoudi, Imen Ben Ismail, Wafa Ragmoun, Hatem Lahdhili, Saber Hachicha, Slim Chenik Clinical Case Reports.2020; 8(12): 2494. CrossRef
- Experience of Penetrating Gunshot Wound on Head in Korea
-
Hong Rye Kim, Seung Je Go, Young Hoon Sul, Jin Bong Ye, Jin Young Lee, Jung Hee Choi, Seoung Myoung Choi, Yook Kim, Su Young Yoon
-
J Trauma Inj. 2018;31(2):82-86. Published online August 31, 2018
-
DOI: https://doi.org/10.20408/jti.2018.31.2.82
-
-
8,025
View
-
73
Download
-
1
Citations
-
Abstract
PDF
Craniocerebral gunshot injuries (CGIs) are extremely seldom happened in Korea because possession of individual firearm is illegal. So, CGIs are rarely encountered by Korean neurosurgeons or Korean trauma surgeons, though in other developing countries or Unites states of America their cases are indefatigably increasing. Management goal should focus on early aggressive, vigorous resuscitation. The treatments consist of immediate life salvage through correction of coagulopathy, intracranial decompression, prevention of infection and preservation of nervous tissue. There have been few studies involving penetrating CGIs in Korea. Here we present a case of penetrating gunshot wound in Korea. We present a 58-year-old man who was unintentionally shot by his colleague with a shotgun. The patients underwent computed tomography (CT) for assessment of intracranial injury. The bullet passed through the left parietal bone and right lateral ventricle and exited through the posterior auricular right temporal bone. After CT scan, he arrested and the cardiopulmonary resuscitation was conducted immediately. But we were unable to resuscitate him. This case report underscores the importance of the initial clinical exam and CT studies along with adequate resuscitation to make the appropriate management decision. Physicians should be familiar with the various injury patterns and imaging findings which are poor prognostic indicators.
-
Summary
-
Citations
Citations to this article as recorded by
- Civilian penetrating traumatic brain injury: A 5-year single-center experience
Omid Yousefi, Pouria Azami, Roham Borazjani, Amin Niakan, Mahnaz Yadollahi, Hosseinali Khalili Surgical Neurology International.2023; 14: 28. CrossRef
|